Mostly Useless Econometrics? Assessing the Causal Effect of Econometric Theory

Abstract

Economics is highly invested in sophisticated mathematics and empirical methodologies. Yet the payoff to these investments in terms of uncontroverted empirical knowledge is much less clear. I argue that leading economics journals err by imposing an unrealistic burden of proof on empirical work: there is an obsession with establishing causal relationships that must be proven beyond the shadow of a doubt. It is far easier to publish theoretical econometrics, an increasingly arid subject that meets the burden of mathematical proof. But the overabundance of econometric theory has not paid off in terms of empirical knowledge, and may paradoxically hinder empirical work by obligating empirical researchers to employ the latest methods that are often difficult to understand and use and fail to address the problems that researchers actually confront. I argue that a change in the professional culture and incentives can help econometrics from losing its empirical relevance. Econometric theory needs to be more empirically motivated and problem-driven. Economics journals should lower the burden of proof for empirical work and raise the burden of proof for econometric theory. Specifically, there should be more room for descriptive empirical work in our journals. It should not be necessary to establish a causal mechanism or a non-parametrically identified structural model that provides an unambiguous explanation of empirical phenomena as a litmus test for publication. On the other hand, journals should increase the burden on econometric theory by requiring more of them to show how the new methods they propose are likely to be used and be useful for generating new empirical knowledge.